amazon

Extract Amazon search result data via a live Chrome session.

5|Updated Apr 9, 2026

Search Tags:#puppeteer #scraping #amazon #data-extraction #web-scraping #cheerio

Ready to run

npx skills add https://github.com/AgentComputerAI/torch --skill amazon

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: amazon
Source: https://github.com/AgentComputerAI/torch/tree/main/skills/sites/amazon
Command: npx skills add https://github.com/AgentComputerAI/torch --skill amazon

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Amazon search pages are server-rendered behind CloudFront with anti-bot walls that block bare curl requests. Attaching to a real Chrome session via the debug port defeats these barriers, enabling reliable data extraction.

Core Features & Use Cases

Real-Chrome scraping: connect to a running Chrome instance via 127.0.0.1:9222 and parse 22 s-search-result cards per page.
Data extraction: pull asin, title, link, price, rating, review count, image, and sponsorship status using Cheerio on the page content.
Use Case: build an affiliate-ready product catalog or price-tracking workflow by saving results to output with a persistent playbook.

Quick Start

Attach to your local Chrome debugging port and start scraping amazon search results.

Frequently Asked Questions about amazon

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I scrape Amazon search results without getting blocked by anti-bot walls?▼

Amazon web scraping requires a real Chrome session attached via the 127.0.0.1:9222 debug port because CloudFront anti-bot walls block bare curl requests. Puppeteer connects to this live Chrome instance to reliably extract 22 search result cards per page.

What data fields can I extract from Amazon search pages using Puppeteer and Cheerio?▼

Using Cheerio to parse HTML from an Amazon search page, you can extract ASIN, title, link, price, rating, review count, image URL, and sponsorship status for each of the 22 result cards captured per page.

Do I need a running Chrome instance to extract data from Amazon search pages?▼

Yes, you need a live Chrome instance running on the local debug port 127.0.0.1:9222. The Skill automates extraction by attaching Puppeteer to this active session to navigate search pages and bypass anti-bot barriers.

Can I scrape multiple pages of Amazon search results with this Puppeteer approach?▼

Yes, the Amazon search scraping approach supports pagination. By leveraging a real Chrome session with Puppeteer, you can navigate through multiple search pages and continuously extract the 22 available result cards from each page.

Why does my web scraping script fail on Amazon search pages while using curl?▼

Amazon search pages are server-rendered behind CloudFront anti-bot walls that actively block bare curl requests. Bypassing these barriers requires attaching Puppeteer to a real Chrome session to mimic genuine browser behavior.